-
Notifications
You must be signed in to change notification settings - Fork 127
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Feature request: graceful malformed packet drop #75
Comments
Good idea, for long runs there are many situations in which there could be a glitch in the data coming in, for example I’ve seen intermittent cases of chunks of data being lost over USB CDC-ACM, especially during context switches when running inside a VM and under situations of high CPU load. |
OK, I thought it was because I had a mismatched firmware/software version, but I checked out v1.9.3 specifically and flashed the v1.9.3 release firmware and I'm still getting this error fairly frequently when I run with
|
Not sure how they're happening, I've never noticed any issues myself with this, though jsmif reported seeing errors with this in issue #75.
Try the latest master branch, hopefully that fixes this particular crash you reported. |
And this should improve general robustness: 87ac1bd |
The error is not currently occurring for me, but I'm also not physically where it was occurring, so I'm not subject to the same background packets. Once I'm back in the same location I'll let you know if it's still occurring (with the old, and then the new, code). Thank you. |
FWIW I left my sniffers (confirmed to be on 1.9.3) running overnight in a hotel last night (so I won't be able to replicate this in the future), and when I got up one of them had got basically the same error as the OP:
Top of the git log for conf:
|
That should be fixed in the latest master branch; it will still work fine with the v1.9.3/v1.9.2 firmware. |
Ah, I didn't pay attention and realize the cited fix commit wasn't in 1.9.3. Thanks |
FYI I don't seem to be seeing any script-killing-crashes due to unpack errors anymore. I do see the occasional error still popping up, but it seems to be handled gracefully, so I'm closing this now until I find another script-killer. Thanks! |
I just went to do a headless long-term sniffing run, and when I got back I saw Sniffle had died prematurely with the following error:
(Note: the device was accidentally out of date, so those line numbers correspond to commit f4cc588).
So the ask would be to wrap instances of struct.unpack(), or anything else which might throw an error on malformed packets in a try-except, to just make it throw away the packet so it doesn't accidentally error out early on long runs. Thank you.
The text was updated successfully, but these errors were encountered: